#ragengineering search results
Vector DBs aren't dead; fast-forward to 2025, and yeah, the "grift" vibe faded as everyone from Postgres (pgvector 0.7's HNSW) to Supabase bolted on vectors like it was an afterthought. Who's stacking what for agentic memory in 2025 #VectorDB #RAGEngineering
Stop using single vector search! Now RAG creates multiple data representations (like topo/transport maps) and lets agents pick the best one. It’s “semantic pointillism” for search. #RAGEngineering #SemanticSearch #MultiRepresentation
🚨 RAG Engineers, brace yourselves: More data ≠ Better results Dumping docs into your RAG system doesn't make it smarter. It can actually decrease accuracy. Yeah, you read that right. More data doesn't always mean better performance. #RAGEngineering #AIAccuracy
Common beginner mistake: Thinking the LLM is the product. It’s not. It’s just one part of the orchestrated system. The edge? → Custom pipelines → Tuned embeddings → Efficient storage + retrieval #ragengineering #AIinfra2025
Vector DBs aren't dead; fast-forward to 2025, and yeah, the "grift" vibe faded as everyone from Postgres (pgvector 0.7's HNSW) to Supabase bolted on vectors like it was an afterthought. Who's stacking what for agentic memory in 2025 #VectorDB #RAGEngineering
Stop using single vector search! Now RAG creates multiple data representations (like topo/transport maps) and lets agents pick the best one. It’s “semantic pointillism” for search. #RAGEngineering #SemanticSearch #MultiRepresentation
Common beginner mistake: Thinking the LLM is the product. It’s not. It’s just one part of the orchestrated system. The edge? → Custom pipelines → Tuned embeddings → Efficient storage + retrieval #ragengineering #AIinfra2025
🚨 RAG Engineers, brace yourselves: More data ≠ Better results Dumping docs into your RAG system doesn't make it smarter. It can actually decrease accuracy. Yeah, you read that right. More data doesn't always mean better performance. #RAGEngineering #AIAccuracy
Vector DBs aren't dead; fast-forward to 2025, and yeah, the "grift" vibe faded as everyone from Postgres (pgvector 0.7's HNSW) to Supabase bolted on vectors like it was an afterthought. Who's stacking what for agentic memory in 2025 #VectorDB #RAGEngineering
Something went wrong.
Something went wrong.
United States Trends
- 1. Nancy Pelosi 92.7K posts
- 2. Ozempic 11.1K posts
- 3. Marshawn Kneeland 54.3K posts
- 4. Paul DePodesta N/A
- 5. Jaidyn 2,631 posts
- 6. Michael Jackson 80K posts
- 7. Sean Dunn 3,533 posts
- 8. RFK Jr 21.2K posts
- 9. Oval Office 32.2K posts
- 10. Subway 44.7K posts
- 11. Sandwich Guy 8,594 posts
- 12. Gordon Findlay 5,220 posts
- 13. Kyrou N/A
- 14. Craig Stammen 2,374 posts
- 15. NOT GUILTY 18.9K posts
- 16. On Melancholy Hill N/A
- 17. #NXXT 1,131 posts
- 18. Kazakhstan 8,921 posts
- 19. Rockies 2,058 posts
- 20. GLP-1 6,826 posts